Estimating glomerular filtration rate in kidney donors: a model constructed with renal volume measurements from donor CT scans.

PURPOSE

To create a model to estimate glomerular filtration rate (GFR) in healthy individuals, such as renal transplant donors, by using renal volume measurements derived from multidetector computed tomographic (CT) scans, serum creatinine level, height, weight, race, and age, and to compare the performance of this kidney volume-based model with the modification of diet in renal disease (MDRD) equation.

MATERIALS AND METHODS

This HIPAA-compliant retrospective study was approved by the institutional review board; informed consent was waived. Age, sex, height, weight, race, serum creatinine level, and measured GFR were recorded from 244 individuals who underwent renal donor evaluation over a 2-year period. An automated segmentation algorithm was used to measure renal parenchymal volume from CT images. GFR was measured by using urinary clearance of iodine 125 ((125)I) iothalamate. Analysis of covariance was used to model GFR measured by using (125)I-iothalamate clearance from the significant variables. The model was tested in 100 different renal donors and performance was compared with performance of the MDRD equation.

RESULTS

Renal volume, age, serum creatinine level, and weight (P < .001) significantly correlated with GFR measured by using (125)I-iothalamate clearance. Sex (P = .6), race (P = .9), and height (P = .76) were not significant. The fitted regression model was GFR(EUn) = 70.77 - 0.444 A + 0.366 W + 0.200 V(R) - 37.317 Cr (r(2) = 0.57), where GFR(EUn) is estimated unadjusted GFR in milliliters per minute, A is age in years, W is weight in kilograms, V(R) is mean total renal volume in milliliters, and Cr is serum creatinine value in milligrams per deciliter (micromoles per liter). Correlation between renal volume-based GFR and GFR measured by using (125)I-iothalamate clearance was +0.42. The model outperformed the MDRD equation in six of six measurements.

CONCLUSION

A model for estimating GFR that incorporates renal volume correlated well with measured GFR and outperformed the MDRD equation in potential living renal donors; this model could be used to estimate donor GFR from CT scans instead of measuring it by using (125)I-iothalamate clearance.